Our objective is to validate a preclinical xenograft model for assessment of the ability of chemical agents to alter surrogate end- point biomarker (SEB) expression in breast carcinoma in situ. We have developed a human breast epithelial cell line, MCF10DCIS.com, which produces ductal carcinoma in situ of the comedo type. Nearly one gram of tissue is obtained 6-8 weeks after injecting 1x10/7 cells in Matrigel into immune deficient mice. We will be able to validate results with the model by direct comparison with results of a clinical study being conducted at the Karmanos Cancer Institute in which patients diagnosed with DCIS are given 100 mg of soy isoflavones daily for three weeks prior to surgery. Surgical specimens are analyzed for expression of a number of markers including connexin 43, p21, bax, bcl-2, p21, and CDK5. However, the preclinical model allows us to treat groups of xenografted mice with chemopreventive agents and determine expression of biomarkers in lesions from treated and untreated at multiple intervals rather than the single time point possible in the clinical study. We will also be able to test additional chemopreventive agents, doses, and combinations. The size of the xenograft lesions will allow us to look at chemopreventive agents, doses, and combinations. The size of the xenograft lesions will allow us to look at many additional biomarkers, both before and after treatment, than the sparse material likely to be available from the clinical specimens. Once validated, MCF10DCIS.com will provide a unique xenograft model of human ductal carcinoma in situ for rapid screening of putative chemopreventive agents. In combination with the MCF10AT1 in situ for rapid screening of putative chemopreventive agents. In combination with the MCF10AT1 xenograft model of progressive premalignant breast disease, the model will also provide a means of validating the relevance of different SEBs for screening chemopreventive agents.